More than 80% of hospice patients experience distressing symptoms that significantly impact comfort in their final days of life. Previous research demonstrated that complimentary therapies such as massage can alleviate symptoms for minutes to hours, but none have examined cumulative effects of massage therapy or cost-effective methods of providing massage to hospice patients. We propose a randomized clinical trial (RCT) comparing usual care versus usual care plus 5 daily massage therapy sessions given by licensed massage therapists (LMT) on symptom outcomes including: a) symptom distress, b) pain, c) mood state and d) sleep quality in a diverse group of 200 hospice patients with cancer. In a pre-test/post-test design with repeated measures preceded by a 5-day run-in period, patients will complete computerized assessments. All patients will receive usual home care level of hospice care, and patients randomized to the experimental group will receive daily massage therapy for 1 week. Outcome scores will be aggregated for each variable from day 2 to day 6 and analyzed using MANOVA, with post-hoc comparisons if significant group results are obtained. Study findings will be used to guide hospice care providers in selecting evidence-based complimentary interventions to relieve distressing symptoms in people at the end of life. A second aim is to determine the costs of the massage intervention and compare the effects of usual hospice care and usual care plus massage on hospice reimbursement and utilization costs. We will use a 2-step activity-based costing methodology for determining the costs of the intervention and the effects on cost outcomes. We hypothesize that the hospice reimbursements and utilization costs will not differ, meaning that the intervention will be cost neutral. A third exploratory aim is to describe the feasibility and the caregiver strain of LMT training 60 family caregivers to give daily massages at home to 60 of the hospice patients who completed Aim 1 procedures. Feasibility will be evaluated by: a) caregiver massage proficiency; b) massage frequency and length given by caregivers; c) caregiver and patient self-report about the massage experience. Strain is defined as the 1-week change in caregiver strain. We expect that family caregivers will be trained to give massages at a proficient level after 2 sessions with training, evaluation and feedback from a LMT. We also expect that caregivers will give 3 daily massages, each for 20 to 30 minutes and report no significant increase in their caregiver strain. Descriptive statistics and content analysis with Atlas.ti will be used to evaluate the feasibility of family caregivers providing massages and to design a clinical effectiveness study using family members as a cost-effective alternative to LMTs.